Multiface 1 Schematic

[deveng]

Does anyone have a schematic for the RME Multiface I or Multiface II?  Specifically the power supply/digital board on the bottom. 

I'm trying to repair a Multiface I that is having boot-up problems.    When it fails to boot it has no LED's lit up, including the red one.  Since I have more than one unit, I've already tried another firewire cable, and a different PCI card.  Its definately something in the Multiface box.  I used a wall wart on the bench to power up and it does light up the red host error LED!  Checked voltages off the switching regulator output caps.  Solid +5V on both boards.  Solid +/- voltages (I.R.C. about +/-14.3V).  So, it appears to be related to the +12V bus power coming in from the PCI card.  Firewire connector looks good.

Its a multi-layer pcb with ground planes on top and bottom so its difficult to visually spot connections for debug.  I have already replaced the "typical" culprits, the through hole electrolytic capacitors.  Still has issues.  I contacted RME and they said  "schematics are not available even for this old models".   

Regards,
Jeff

 

[Khron]

How did you determine that "it appears to be related to the +12V bus power coming in from the PCI card"? Is there 12V being supplied through the connector / cable from the PCI card to the Multiface?

Some photos might help... Especially since not many of us (i'd wager) have seen inside those units.

 

[deveng]

Yes the PCI card supplies 12v bus power through the 1394 cable.  The unit can also be powered by a wall-wart.  So for bench work I use the wall wart.  Since it always powers up with the wall wart on the bench and I've used other PCI cards to verify its not a lack of 12v bus power getting to the unit, I'm assuming its not getting the bus power past the 1394 connector or there are failed components past the 1394 input in the Multiface.

My next step will be to remove the pcbs from the Multiface, connect to 1394 for 12v bus power.  Then measure pin 1 of the connector to see if its at least there at the pcb.  From there its hard to tell and a schematic would be helpful!


Regards,
Jeff

 

[Khron]

So i'm guessing that photo wasn't taken by you, then?

But yes, starting at checking if the voltage does make it onto the board is a logical first step, indeed. My guess would be all those big fat diodes in the corner there have to do with "steering" the wall-wart / cable power.

LM2585 - http://www.ti.com/lit/ds/snvs120g/snvs120g.pdf
LM2595 - http://www.ti.com/lit/ds/symlink/lm2595.pdf

Whichever input supply source is connected (although i'm guessing the wallwart might have "priority" somehow; or at least it should), it will have to make its way to the respective input pins of those two bigger switching regulators.

Can't make out the markings on those two smaller SOT-223-packaged ones, but my guess is they're "downstream" of the others (likely hanging off the 5V output, regulating that down to 3.3V and 2.5V / 1.8V respectively).

What's under that silver can / heatsink down in the corner?

 

[Tubetec]

Just by IC21 ,is that signs of overheat by the pins ? must be a three terminal regulator .

 

[deveng]

The picture was not mine.  Just grabbed it from the net.  So any signs of heat was someone else's problem, hah! My pcb's look pristine.  Tubetec, thanks for pointing out the 3 terminal regulators as I have not checked their outputs.

I will check those diodes with bus power attached.  When on the bench with wall wart power, the two switching regulators appeared to be working fine and stable.  I don't remember exactly what is under the silver shield but I believe it was related to the headphone output. 

Khron,  good point about the possibility of other regulators for "3.3V and 2.5V / 1.8V".  The Xilinx Spartan FPGA will run off 3.3v or lower.  I had not thought about that.  I'll check those SOT-223 devices.  I would expect it controls the LED on/off status so if it fails to get power, or too low, no lights.


Regards,
Jeff

 

[deveng]

Pulled the pcb's and checked the 3 terminal regulator, good 3.6 volts.  Connected to firewire/1394 cable and booted up with PCI bus power and its been running perfect with no disconnect.  This kind of intermittent issue just sounds like a connector issue to me.  My first close inspection didn't reveal anything, but after looking again at pin 1 of the firewire (bus power) with a loop, it looks a bit fishy.  I'll re-flow the 1394 connector to the pcb and see what happens.  Hopefully, my solution is that easy!

regards,
Jeff